Considering ethical issues raised by AI programs applied to people's medical records, or to driving cars.

The ethical issues mentioned in this article are not the biggest ones that these technologies raise.

For medical records (including prescriptions), the biggest issue is not whether IBM's program should be allowed to chew on them, but a logically prior issue: whether everyone's medical records should be put in a centralized data base where the state can trivially look at them, even in less sophisticated ways.

For driverless cars, assuming are as safe as human drivers, who to blame for the occasional accident will be a side issue. The biggest issues will be whether they construct for each person a travel dossier that GCHQ or the NSA can study years later, and whether they condemn today's millions of paid drivers to the underclass of the permanently unemployed.

In the times of increasing unemployment, and redistribution of wealth to the rich, efficiency is a misguided goal.